Accurate modeling of the structure and recognition of adaptive immune receptors is a major challenge in computational biology. Despite a shared immunoglobulin structural framework, highly variable antigen binding loop sequences and structures, with intrinsic dynamics and binding conformational changes, are often not accurately represented or correctly modeled using current algorithms. There is an even greater need to address this challenge due to the rapidly growing field of immune sequencing, which often results in thousands of sequences of antigen-specific immune receptors from the repertoire of a single individual per experiment. In the absence of reliable modeling tools, the observed shared sequence motifs and areas of divergence lack a structural and mechanistic explanation, given that experimental structural characterization is not practical or feasible for more than a handful of molecules. The focus of this application is on T cell receptors (TCRs), which recognize antigenic peptides by the major histocompatibility complex (MHC), leading to the cellular immune response. We will develop advanced modeling and design algorithms to address the challenges of flexible loop modeling through informatics and knowledge-based developments to help unravel their recognition code. This will entail the development of algorithms to reliably model TCR structures from sequence (Aim 1), model TCR recognition of peptide-MHCs through docking (Aim 2), and design TCR recognition through loop engineering (Aim 3). These Aims will be accomplished through validation against existing experimental structural and affinity data, as well as close partnership with experimental laboratories that will provide sequence, structural, dynamic, and binding measurements of TCRs, and validate affinity and structure of designed receptors. Collectively, these developments will allow the illumination of the mechanistics underpinning recognition by specific and repertoire-level TCRs from sequence, improved loop modeling and docking algorithms, and the capability to effectively control and engineer TCR recognition through structure-based design.